LUCI Desktop is a fully local memory assistant for Mac and Windows that indexes screen activity and calls to supply context to AI coding agents.
What the product does and how it is positioned
LUCI Desktop is an on-device personal memory tool designed to capture and index user activity across desktop applications. It records on-screen content and meeting audio from tools like Zoom, Google Meet, and Microsoft Teams without requiring application-specific connectors or cloud accounts.
The software operates fully locally, utilizing on-device language models like Phi, Qwen, Llama, and DeepSeek while encrypting all stored data on disk. It connects to developer tools such as Cursor, Claude Code, and Codex, allowing AI agents to query past visual context, meeting transcripts, and project activities directly from local plain files.
Source-supported ways to use the product
Developers use LUCI to let AI coding agents like Cursor or Claude Code recall terminal errors, design specifications, and previous discussions without manual copy-pasting.
Team members record calls across Google Meet, Zoom, and Teams to retain searchable transcripts and generate meeting notes automatically.
Founders and professionals review automated daily summaries stored in a local Life folder to track where time was spent across projects and applications.
Knowledge workers query past screen views in natural language to locate previously viewed references, documents, and discussions.
LUCI Desktop provides a background indexing pipeline that monitors screen state and voice interactions directly on the host computer. By processing visual information and audio locally, it maintains a continuous chronological timeline without relaying desktop data to external servers or requiring OAuth integrations for individual apps.
The platform integrates with developer agents such as Codex, Cursor, and Claude Code through an agent bridge. External agents query LUCI's local indices to retrieve relevant project constraints, architectural decisions, and visual references from earlier sessions.
Data governance relies on client-side boundaries. Sensitive items such as API credentials, authentication tokens, and credit card numbers are redacted prior to persistence. Data remains encrypted on disk, and summaries are saved as plain text files that users can inspect, retain, or delete on custom schedules.
Checks to run with your own material and workflow
What was checked and when
Answers based on the source-checked product record
LUCI Desktop is documented for Mac and Windows operating systems, running locally on the user's computer.
No, LUCI operates fully locally, storing encrypted records on disk without creating cloud copies of user data.
LUCI documents integration with agents including Claude Code, Cursor, Codex, Gemini, Copilot, Windsurf, Zed, and VS Code.
LUCI includes automated zero-leak redaction that blacks out payment cards, access keys, and passwords before data is written to disk.
The platform supports running local models such as Phi, Qwen, Llama, and DeepSeek on the host machine.
